4 tips for user activation!

Today, let’s talk about how to define user activation and teach you 4 tips to quickly define user activation behavior for your product.

Step 1: Propose a behavior that may lead to user activation

This is the open exploration phase of defining users. The main purpose is to clarify the long-term value of the product and find the fastest way for new users to feel the long-term value when they start using the product.

Then, according to the above method, several most likely new user activation behaviors are proposed.

Usually, you can use the following two methods to preliminarily list several potential user activation behaviors.

Method 1: Through key questions

• WHO Who are the users?
• WHAT What problem do users need to solve with this product?
• WHY Should users solve this problem?
• What other ways are there for VS users to solve this problem?

The general usage scenario is to find the long-term value of the product and the behaviors required to experience these values ​​through key questions, and then infer what behaviors new users can complete in the short term.

Taking the beauty camera as an example, let’s apply the formula.

• who: female
• what: take beautiful photos
• Why: Let friends see your beauty
• vs: Take the photo first, then edit it

Possible actions to activate:

• Photograph
• Use filters
• Share to Moments

Which ones can be completed quickly?

The first 2.

OK, then we can make a preliminary judgment on the possible user activation behaviors when taking a photo and saving it.

Method 2: Through user research

Simply put, it is to compare different user answers, discover the most important value of the product to users and find alternative activation behaviors.

You can conduct surveys on the following different user roles.

• Long-term most active users: Why do they find the product valuable?
• Users who leave quickly after registering: Why do they leave?
• Active users after registration: Why do they stay?

The usage scenario is generally that if there are a large number of alternative behaviors, user research can help narrow the number of alternative behaviors. This is especially important for products with multiple usage scenarios and functions.

Tip 2: Find the most critical user activation behavior

Here are two steps to share with you.

Step 1: Find the activation period for new users and assess how quickly activation occurs.

Step 2: Compare the early retention curves to find the behavior that has the greatest impact on early retention during the activation period of new users.

How to find out how long?

Principle 1: The higher the frequency of use, the faster the activation needs to be

The higher the frequency of use, the sooner new users expect to obtain value from the product. The activation period of new users can be roughly determined based on the frequency of use.

Principle 2: The shorter the life cycle, the faster the activation needs to be

The shorter the life cycle, the sooner new users expect to get value from the product

Principle 3: Refer to actual data

Analyze the actual data of new users and look at the real-time window of most early activation behaviors

Example: Pull the time distribution of all users who have potential activation actions for the first time, and judge based on more than 80% of the behaviors

Assumption: Taking Beauty Camera as an example, assuming that Beauty Camera confirms the first day as the new user activation period, the corresponding earliest retention is the first 31 days.

So, how do you compare retention curves manually?

1) Collect retention data of new users in the first 31 days

2) Group users by whether they have a certain behavior and collect retention data

1. All new users
2. First day using filters
3. Day 1: No filter
4. Day 1 Photoshoot
5. No photos on the first day

3) Draw the first 31-day retention curves for different user groups

4) Compare the retention curves to find the biggest difference in retention with and without the behavior

The bigger the gap, the more likely it is to be an Aha moment.

Tip 3: Calculate the magic number

What is the magic number?

The optimal number of key rows, also known as the magic number, is found through data analysis.

However, not every magic number is the same for every product.

For some activation actions, it is enough to do it only once, such as payment collection in e-commerce.

Some activation behaviors need to be repeated multiple times to ensure that users feel the value, such as watching short videos.

Theoretically, the more repetitions, the greater the improvement in retention, but activation time for new users is a priority, so it is unrealistic to ask users to repeat multiple times.

Therefore, we want to find the optimal number of activation times to ensure that users get value while not burdening them.

Here I can briefly introduce a commonly used method, called the maximum marginal utility method.

1. Draw a distribution chart of the number of activation behaviors of new users on the first day
2. Analyze the relationship between the number of first activation behaviors and retention rate
3. Find the number of activation behaviors corresponding to the point with the largest marginal benefit of retention (the inflection point of retention rate, that is, the number of times with the largest marginal utility)

Let’s take the example of the number of times the beauty camera uses filters.

From the inflection point above, we can preliminarily conclude that the user’s activation behavior can be defined as the first time a new user uses a filter.

Step 4: Test to verify causality

The first three steps are, in fact, our initial preliminary estimates after all, and they cannot prove that they are definitely user activation behaviors. Here, the previous behaviors can be regarded as related behaviors.

If you really want to anchor a user activation behavior, you also need to verify causality through A/B testing to confirm that pushing users to complete early key behaviors can indeed improve retention.

So what is causal behavior? How is it different from related behavior?

To explain briefly: correlation generally refers to the observation that a certain early behavior is accompanied by higher retention rates. Causality means that the user took some early action, which led to a higher retention rate.

For example, rainy days and holding an umbrella are causally related, and rainy days and wet ground are also causally related. However, holding an umbrella and wet ground are not causally related, but only correlated.

Here I take the example of Kugou, which I love to use. Kugou found that new users who sing have better retention, and believe that user activation behavior is singing, but it is necessary to verify causality and confirm whether the magic number is 1 or 3 songs.

Experiment: A/B testing the new user guide page

Control group: no singing

Experimental group 1: sing 3 songs

Experimental group 2: sing 1 song

Results: Experimental group 2 had the best results, with a retention rate of 10% the next day.

Only in this way can we finally determine that this key behavior can be defined as user activation.

Finally, let’s talk about my thoughts.

We always talk about activation and retention. In fact, they are not isolated relationships, but corresponding relationships. When we define activated users, we often need to use retention data.

Therefore, if we want to improve the retention rate, we also need to trace back to the user's activation behavior, so that they correspond to each other and are connected in series.

Author: Lei Zhenzi said

Source: Lei Zhenzi said